AI for Science

Manually curated biomedical repositories -- spanning bioactivity, genomics, and chemistry -- are expensive to maintain, lag behind primary literature, and discard experimental context, obscuring nuances needed to assess data correctness and coverage. We show that PubMed itself can be autonomously and cost-effectively turned into structured datasets that are larger, more nuanced, and more accurate than the curated databases they replace. We present three coupled contributions: (1) an LLM-based entity-tagging pipeline, grounded in nine biomedical ontologies, that tags 4.5B entities across 19 categories in a 22.5M-paper, 2.5T-token PubMed corpus; (2) hybrid sparse-dense retrieval supporting entity-filtered semantic queries over the tagged corpus; and (3) Starling, a multi-agent deep research system that, given only a natural-language task description, designs precision- and recall-targeted retrieval filters, induces an extraction schema, and emits structured records with nuance-rich fields and supporting passages. Across six tasks -- blood-brain barrier permeability, oral bioavailability, acute toxicity (LD50), gene-disease associations, protein subcellular localization, and chemical reactions -- Starling produces ~6.3M records (91K-3M per task); several are, to our knowledge, the largest public datasets for their property. Frontier-model rejection of our extractions is 0.6-7.7% across tasks, far below error rates we measure on widely used curated counterparts (e.g., 16.5% on BBB_Martins, 7.3% on Bioavailability_Ma). Beyond scale and accuracy, the supporting passages carry nuance tabular databases discard -- e.g., oral bioavailability may depend on fed vs. fasted state. Together, the corpus, retrieval, and agent establish a foundation for AI-driven therapeutic design. Code and datasets: https://github.com/starling-labs/starling.

Water-based Liquid Scintillator (WbLS) was proposed over a decade ago as a novel detector medium that might allow the separation and tuning of the relative ratio of the Cherenkov and Scintillation signals. A detector deploying this technology could combine GeV-scale and MeV-scale neutrino detection at scale. Furthermore, the metal-loading capability of such a material enables neutron tagging and allows the effective particle containment to be tuned. WbLS is attractive both for the potential to use it in large detectors and the ability to modify the configuration in situ. At Brookhaven National Laboratory (BNL), two prototypes have been built for understanding WbLS properties and stability, with masses of 1-ton and 30-ton, respectively. We present here the 30-ton prototype detector design, installation, and operation. Results from the analysis of data collected in the two detectors will follow in future publications.

Federated learning (FL) has emerged as a promising paradigm for managing electric vehicle (EV) battery data in intelligent transportation systems (ITS), enabling privacy-preserving tasks such as anomaly detection and capacity estimation. However, most existing frameworks rely on centralized aggregation schemes, which pose critical limitations in terms of security and trust. To address these challenges, we propose ABC-DFL, an automated Byzantine-resilient clustered decentralized federated learning (C-DFL) framework for connected EVs. The proposed incentive-driven C-DFL system replaces the central server with an open-permissioned blockchain, featuring a new dynamic Quorum Byzantine Fault Tolerance (QBFT) protocol and an oracle-based aggregation layer, to enhance trust, security, and automation. At the core of ABC-DFL lies FLECA (Filtered Layered Enhanced Clustering Aggregation), a robust hierarchical aggregation protocol that mitigates Byzantine attacks by having each EV filter malicious updates using an adaptive threshold based on deviations from its reference model update. Oracle nodes, responsible for inter-group aggregation, employ robust clustering to isolate and aggregate model updates from trustworthy EV groups. Comprehensive experimental evaluations demonstrate that FLECA matches FedProx convergence under benign conditions and significantly outperforms existing defenses with attack impact scores below 0.10 in adaptive adversarial scenarios. Furthermore, several learning experiments with multitask models confirm the effectiveness and fairness of the incentive mechanism. Finally, on-chain and off-chain benchmarks validate the practicality of ABC-DFL.

Small and Medium Enterprises (SMEs) constitute 99.9% of U.S. businesses and generate 44% of economic activity, yet systematically identifying high-potential SMEs remains an open challenge. We introduce SME-HGT, a Heterogeneous Graph Transformer framework that predicts which SBIR Phase I awardees will advance to Phase II funding using exclusively public data. We construct a heterogeneous graph with 32,268 company nodes, 124 research topic nodes, and 13 government agency nodes connected by approximately 99,000 edges across three semantic relation types. SME-HGT achieves an AUPRC of 0.621 0.003 on a temporally-split test set, outperforming an MLP baseline (0.590 0.002) and R-GCN (0.608 0.013) across five random seeds. At a screening depth of 100 companies, SME-HGT attains 89.6% precision with a 2.14 lift over random selection. Our temporal evaluation protocol prevents information leakage, and our reliance on public data ensures reproducibility. These results demonstrate that relational structure among firms, research topics, and funding agencies provides meaningful signal for SME potential assessment, with implications for policymakers and early-stage investors.

We simplify and improve the main fundamental theorems of positive characteristic generic vanishing theory. As a quick corollary of the theory, we prove that a normal variety $X$ of maximal Albanese dimension satisfies $H^0(X, ω_X) \neq 0$ and that if $\mathrm{Alb}(X)$ is ordinary, then $S^0(X, ω_X) \neq 0$.

We prove that for $n \leq 4$ and $p > 5$, quasi--Gorenstein $F$--pure and $\mathbb{Q}_p$--rational $n$--fold singularities are canonical. This is analogous to the usual fact that rational Gorenstein singularities are canonical. The proof is based on a careful analysis of the dual complex of a dlt modification of a log canonical singularity. The result for $n = 4$ is contingent upon the existence of log resolutions.

Let $W_d(n)$ be the number of $2n$-step walks in $\mathbb{Z}^d$ which begin and end at the origin. We study the exponent of $2$ in the prime factorisation of this number; i.e., $w_d(n) = ν_2(W_d(n))$. We show that, for each $d$, there is a relationship between $w_d(n)$ and the number $s_2(n)$ of $1$s in the binary expansion of $n$. For example, $w_d(n) = s_2(n)$ if $d$ is odd and $w_d(n) = 2s_2(n)$ if $ν_2(d) = 1$; while $w_d(n) \ge 3s_2(n)$ if $ν_2(d) = 2$. The pattern changes further when $ν_2(d) \ge 3$. However, for each $d$, we give the best analogous estimate of $w_d(n)$ together with a description of all $n$ where equality is attained. The methods we develop apply to a wider range of problems as well, and so might be of independent interest.

This paper classifies complex local representations of the twisted virtual braid group, $TVB_2$, into $\mathrm{GL}_3(\mathbb{C})$. It shows that such representations fall into eight types, all of which are unfaithful and reducible to a two-dimensional representation. Further reducibility to a one-dimensional representation is analyzed for specific types. The paper also examines complex homogeneous local representations of $TVB_n$ into $\mathrm{GL}_{n+1}(\mathbb{C})$ for $n \geq 3$, identifying seven unfaithful types. Additionally, complex local representations of the singular twisted virtual braid group, $STVB_2$, into $\mathrm{M}_3(\mathbb{C})$ are classified into thirteen unfaithful types. Finally, the paper demonstrates that not all complex local extensions of $TVB_2$ representations to $STVB_2$ conform to a $Φ$-type extension.

The instrumental variable (IV) approach is commonly used to infer causal effects in the presence of unmeasured confounding. Existing methods typically aim to estimate the mean causal effects, whereas a few other methods focus on quantile treatment effects. The aim of this work is to estimate the entire interventional distribution. We propose a method called Distributional Instrumental Variable (DIV), which uses generative modelling in a nonlinear IV setting. We establish identifiability of the interventional distribution under general assumptions and demonstrate an 'under-identified' case, where DIV can identify the causal effects while two-step least squares fails to. Our empirical results show that the DIV method performs well for a broad range of simulated data, exhibiting advantages over existing IV approaches in terms of the identifiability and estimation error of the mean or quantile treatment effects. Furthermore, we apply DIV to an economic data set to examine the causal relation between institutional quality and economic development and our results align well with the original study. We also apply DIV to a single-cell data set, where we study the generalizability and stability in predicting gene expression under unseen interventions. The software implementations of DIV are available in R and Python.

Large language models can write scientific code, but direct paper-to-program translation remains fragile when correctness depends on tacit conventions rather than explicit equations. We frame this as a \textbf{knowledge-externalization} problem: index choices, gauges, fermionic signs, contraction order, validation gates, and scaling constraints must be made explicit before code generation. We evaluate a multi-stage, human-in-the-loop workflow on two quantum many-body tasks. DMRG from Schollwock's pedagogical review serves as calibration: specification-guided implementations pass in all 16 model pairings, compared with 6/13 direct attempts, and a prose-specification ablation shows that externalized content, not \LaTeX{} form, is the active ingredient. Pfaffian conversion of HFB states to MPS from the five-page Letter by Jin et al. serves as the stress test: no public implementation is available, and success depends on tacit sign, gauge, ordering, and scalability conventions. Here the workflow yields 11/26 audited passes, while direct prompting yields none. Cross-specification transfer is asymmetric: non-GPT specifications implemented by GPT~5.5 pass 4/4, whereas GPT~5.5 specifications implemented by weaker models fail 4/4. The contrast supports a two-bottleneck picture. Externalization resolves the first bottleneck -- paper-to-code ambiguity -- well enough to make DMRG reproducible and Pfaffian-MPS auditable. The remaining failures expose a second bottleneck in implementation-model capability. Iterative meta-specification moves this boundary but does not eliminate it. The resulting \emph{Paper-to-Program Many-Body} skill is both a reusable implementation protocol and a diagnostic instrument for AI-assisted many-body programming.

We prove that the Hirzebruch quadratic form of a complex hyperplane arrangement is non-positive on the set of stable weights, and we identify the zero locus within this set with flat logarithmic connections of a distinguished type. The proof uses Kempf--Ness and the frame-potential inequality.

Accurate epidemic forecasting requires models that account for the layered and heterogeneous nature of real social interactions. The basic reproduction number $\mathcal R_0$, as calculated from models that assume homogeneous mixing or single-layer contact structures, has limited applicability to complex social systems. Here, we derive an expression for $\mathcal R_0$ in the context of multiplex networks, enabling the analysis of disease transmission across multiple social layers. We adapt the Degree-Based Mean-Field (DBMF) SIR model for single-layer complex networks to the multiplex setting, where each layer is characterized by its own degree distribution and infection rate. Using the Next Generation Matrix method, we derive an analytical expression for the basic reproduction number $\mathcal R_0$. Numerical integration of the multiplex DBMF equations shows that $\mathcal R_0=1$ marks the epidemic threshold and governs the behavior of key outbreak indicators as expected. In addition to the exact expression for $\mathcal R_0$, we introduce an approximation, denoted by $τ$, which is simpler to compute and admits a more transparent interpretation in terms of the epidemiological and topological parameters of the system. Stochastic agent-based simulations support these findings, demonstrating a direct correspondence between $τ$ and the average number of secondary infections generated during the early stages of an outbreak, consistent with the epidemiological interpretation of $\mathcal R_0$. This work provides a robust generalization of $\mathcal R_0$ for layered contact structures, offering a more realistic basis for epidemic forecasting and the design of intervention strategies.

This paper studies the clustering behavior of weakly interacting diffusions under the influence of sufficiently localized attractive interaction potentials on the one-dimensional torus. We describe how this clustering behavior is closely related to the presence of discontinuous phase transitions in the mean-field PDE. For local attractive interactions, we employ a new variant of the strict Riesz rearrangement inequality to prove that all global minimizers of the free energy are either uniform or single-cluster states, in the sense that they are symmetrically decreasing. We analyze different timescales for the particle system and the mean-field (McKean-Vlasov) PDE, arguing that while the particle system can exhibit coarsening by both coalescence and diffusive mass exchange between clusters, the clusters in the mean-field PDE are unable to move and coarsening occurs via the mass exchange of clusters. By introducing a new model for this mass exchange, we argue that the PDE exhibits dynamical metastability. We conclude by presenting careful numerical experiments that demonstrate the validity of our model.

We study a diffuse--interface model for a binary incompressible mixture in a periodically perforated porous medium, described by a time-dependent Navier--Stokes--Cahn--Hilliard (NSCH) system posed on the pore domain $Ω_p^\varepsilon\subset\mathbb{R}^3$. The microscopic model involves a variable viscosity tensor, a non-conservative source term in the Cahn--Hilliard equation, and mixed boundary conditions: no-slip on the outer boundary and Navier slip with zero tangential stress on the surfaces of the solid inclusions. The capillarity strength $λ^\varepsilon>0$ depends on the microscopic scale $\varepsilon>0$. The analysis consists of two main parts. First, for each fixed $\varepsilon>0$ we prove existence of a weak solution on a finite time interval $(0,T)$ and derive a priori estimates that are uniform with respect to $\varepsilon$ (and $λ^\varepsilon$). Second, we perform the periodic homogenization for the perforated setting in the limit $\varepsilon\to0$. Depending on the limit value $λ$ of the capillarity strength $λ^\varepsilon$, we obtain two distinct effective models: (i) in the vanishing capillarity regime $λ=0$, the limit system decouples completely into a standalone linear Stokes system for the velocity--pressure pair and a standalone Cahn--Hilliard system with source term $G$ for the phase field and chemical potential, with no macroscopic convection, advection, or capillary coupling between the two; (ii) in the balanced regime $λ\in(0,+\infty)$, we derive a Navier--Stokes--Cahn--Hilliard system with nonlinear convection and advective transport of the phase field at the macroscopic scale, coupled through a capillary forcing term. Finally, we establish the convergence of the microscopic free energy to a homogenized energy functional satisfying an analogous dissipation law.

The deep neural network is a widely used framework in machine learning that has been widely applied in various fields. However, deep neural networks often involve a large number of parameters and inputs, many of which may be irrelevant to the goal or true output. These parameters and input variables not only increase computational complexity, but also contribute to additional computational cost. One solution to this problem is knockoff methods, which have proven successful in controlling false discovery rates in high-dimensional regression. Building on the knockoff methods and using the regularised neural network, this paper proposes three variable screening methods under the condition of controlling false discovery rates: one layer filter, multiple layers filter, and variable weight aggregation filter. In comparison with existing algorithms, we find that our algorithms show satisfactory performance.

Diffusion-based neural solvers have shown strong promise for combinatorial optimization (CO), but existing methods typically rely on supervised training with large collections of near-optimal solutions. In this work, we extend adjoint-based trajectory optimization methods to discrete combinatorial domains. We formulate diffusion-based CO as a stochastic control problem over Continuous-Time Markov Chains and introduce discrete adjoint dynamics for propagating optimization signals through discrete generative trajectories. Building on this formulation, we propose Combinatorial Adjoint Matching (CAM), an unsupervised training framework for discrete diffusion solvers with structured and low-variance trajectory-level optimization signals. Empirically, CAM consistently outperforms existing unsupervised diffusion baselines and achieves performance competitive with strong supervised diffusion solvers and even traditional solvers across diverse combinatorial optimization problems. Our code is available at https://github.com/Shengyu-Feng/CAM.

Cryptocurrencies are increasingly adopted as investment assets, making their interactions with traditional financial markets central to cross-asset diversification and systemic risk. This paper studies the integration of cryptocurrencies, fiat currencies, and S&P500 equities using a balanced panel of 381 assets from October 2017 to February 2024. We combine rolling correlation networks, community structure, market-specific and system-wide Turbulence Indices, and VAR-based connectedness analysis to examine how market stress, network structure, and shock transmission vary across financial regimes. The results show that cross-asset integration is episodic. In calm periods, the three asset classes remain relatively segmented, whereas under stress, local clustering increases, modular separation weakens, and communities become more compositionally mixed across asset classes. Connectedness analysis further shows that regime shifts alter the structure of transmission rather than simply increasing spillover magnitudes. In high-turbulence states, fiat-market turbulence becomes the dominant propagation channel, while network clustering and modularity play a greater role in transmitting forecast uncertainty. These findings support the interpretation of network structure as an emergent, state-dependent transmission layer rather than a persistent exogenous driver of turbulence. The results highlight the need for regime-aware risk monitoring, since full-sample connectedness estimates can understate the cross-asset coupling that emerges precisely when diversification benefits are most fragile.

Schrödinger bridges for time series (SBTS) generate synthetic paths by projecting, in relative entropy, a Brownian reference onto the path laws that match the joint distribution of the data on the observation grid. The Brownian reference, however, fixes the quadratic variation of the generated paths, which is restrictive when stochastic volatility, correlated noise, or rank-deficient covariance structures must be reproduced. We introduce "Triangular-Reference Schrödinger Bridges for Time Series" (TR-SBTS), which keeps the entropy-projection backbone of SBTS but replaces the Brownian reference by a triangular, volatility-informed, intervalwise frozen reference on a state augmented with latent covariance descriptors. The construction remains a single entropy projection on the augmented state: the minimiser is the \(h\)-transform of the reference, and on each frozen interval the optimal drift has the logarithmic-gradient form \(b^\star(t,x)=A\,\nabla\log H(t,x)\), intrinsic to the active covariance directions when the frozen covariance \(A\) is degenerate. We prove stability of the frozen approximation and consistency of the associated regularised kernel estimators, describe a reference-aware Nadaraya--Watson implementation of the conditional next-increment law, and evaluate the construction on numerical experiments.

We introduce the \emph{Morse ensemble polynomial} $\ME_K(z_0,\ldots,z_d)$ of a finite simplicial complex $K$, defined as the generating function $\ME_K = \sum_M \prod_i z_i^{c_i(M)}$ over all acyclic matchings $M$ on the face poset of $K$, where $c_i(M)$ counts critical $i$-simplices. This polynomial records the complete distribution of Morse vectors across all discrete Morse functions on $K$, and is an isomorphism invariant of simplicial complexes. Our main results are the following. \textbf{(I) The Laplacian Formula}: for any connected graph $G$, $\ME_G = z_1^{m-n}\det(z_0z_1\,I_n + L_G)$, identifying $\ME_G$ as a complete Laplacian spectral invariant and showing $\ME_G$ to be incomparable with the Tutte polynomial. \textbf{(II) The Top-Face Recursion}: adding a $d$-simplex $σ$ (with $\partialσ\subset K$) to a complex $K$ gives a recursion $\ME_{K\cup\{σ\}} = z_d\cdot\ME_K + \sum_{τ\precσ}(\ME_{P(K')\setminus\{σ,τ\}}-F(K,σ,τ))$. The correction term is controlled by the top incidence graph: an incidence-separation criterion detects exactly when $F=0$, and the incidence distance gives the leading obstruction term. As a topological application, this recursion gives exact coefficient recursions for perfect and optimal discrete Morse vectors. \textbf{(III) The independence ME polynomial} $Φ(G) := \ME_{\mathrm{Ind}(G)}$ is a fine graph invariant which strictly refines the graph-level Morse ensemble $\ME_G$, separates examples not distinguished by $T_G$ and $I(G;t)$, and records collapse-level information of $\mathrm{Ind}(G)$ through coefficients such as $[z_0]Φ(G)$.

We define a sequence of integer-valued invariants $γ^k(L)$ for a $3$-component link $L$. We prove that the resulting $γ$-invariants are invariant under concordance, and more generally under weak cobordism, and that they lift certain Milnor invariants of 3-component links. To establish this, we introduce an invariant $h(L)$, a $3$-component analogue of the Kojima--Yamasaki $η$-invariant, and show that it recovers the $γ$-invariants. As applications, we obtain a weak-cobordism classification when the distinguished component has trivial Alexander polynomial and characterize knots that bound continuously embedded disks in $B^4$ whose complements have fundamental group $\mathbb{Z}$.

We develop a first and second fundamental theorem for $n$--tuples of bosonic and fermionic matrices, by developing graded analogues of the classical case.

The exodromy correspondence of Barwick, Glasman, and Haine computes constructible sheaves of spaces on a scheme $X$ as an $\infty$-category of continuous functors from the profinite category $\operatorname{Gal}(X)$. Viewing $\operatorname{Gal}(X)$ instead as a condensed category, this was extended by Wolf to an exodromy correspondence for pro-étale sheaves. Using the condensed perspective from the outset, we give a quick and self-contained proof of the pro-étale exodromy theorem. This is used to extract an exodromy theorem for (Postnikov complete) étale sheaves that does not yet appear in the literature, which is closely related to Lurie's work on ultracategories. Finally, we use this to give a new proof of the constructible étale exodromy correspondence of Barwick, Glasman, and Haine. Without additional effort, our method removes the qcqs hypotheses on the schemes, and gives versions for sheaves with coefficients in more general $\infty$-categories. Finally, we refine the methods to obtain a $κ$-condensed statement for any uncountable cardinal $κ$ such that $κ> \lvert \mathcal O_X(U) \rvert$ for every affine open $U \subseteq X$.

Already in $\bf{R}^4$, there are many minimal hypersurfaces, yet few structural results. We show that minimal submanifolds, of any dimension and codimension, that are confined in space are very restricted. It is well-known that the half-space theorem fails already for hypersurfaces in $\bf{R}^4$, where there are many examples contained in a slab. In $\bf{R}^3$ the height of the catenoid grows at a logarithmic rate, whereas in higher dimensions the height of the catenoid remains bounded. We will see that even in high dimensions, minimal submanifolds that are confined in space must satisfy strong structural restrictions. We show that any proper minimal immersion whose height grows sublinearly must have Euclidean volume growth. A consequence is an optimal Bernstein theorem in any dimension for stable hypersurfaces with sublinearly growing height that generalizes results of Moser, Bombieri-De Giorgi-Miranda, Trudinger, Caffarelli-Nirenberg-Spruck and Ecker-Huisken. Euclidean volume growth is a powerful property and there are many other consequences.

Small area estimation produces estimates of population parameters for geographic and demographic subgroups with limited sample sizes. Such estimates are critical for policy decisions, yet principled validation of these models remains a challenge. Unlike conventional predictive settings, validation data are rarely available. Data thinning splits a single observation into independent training and test components. It enables out-of-sample validation using only the area-level summary statistics routinely available, requiring only their Gaussianity and known sampling variances. However, the properties of thinning-based model comparison have not been formally studied. In this paper, we develop these properties. We construct an unbiased estimator of thinned-data mean squared error and show that it differs systematically from its full-data counterpart; for the standard Fay-Herriot model, the gap admits a closed-form expression that depends on the candidate model's shrinkage behavior. We further show that the estimator variance increases sharply as the training fraction approaches one, producing a bias-variance tradeoff with no universally optimal thinning parameter. Practical recommendations balancing these forces are informed by theory and verified empirically. Design-based simulations using American Community Survey microdata show that the recommended data thinning approach is competitive with information-criterion and simulation-based methods, and substantially more stable across heterogeneous sampling designs.

We take two approaches to classifying the complexity of Presburger models: Scott analysis and degree spectra. In particular, we investigate the possible Scott sentence complexities and possible degree spectra of models of Presburger arithmetic. Many of our results will be achieved by showing how given a linear order $\mathcal{L}$, we can construct a Presburger group $P_\mathcal{L}$ that maintains much of the structure of $\mathcal{L}$.

We consider an interacting particle system, which generalizes the classical totally asymmetric simple exclusion process (TASEP), in that each site can contain up to a fixed finite number of particles, and the particle movement is governed by a {\em back-pressure} (BP) algorithm (also often called {\em MaxWeight}). There are $N$ sites (with $N$ finite or infinite), each may contain at most $c$ particles, $1 \le c < \infty$. New particles enter the system at the left-most site $1$ as a Poisson process of rate $α\le 1$, unless site $1$ has $c$ particles. Particles (if any) are removed from the right-most site $N$ as a Poisson process of rate $β\le 1$. The left-to-right movement of particles between neighboring sites is governed by the BP rule: one particle moves from site $n$ to $n+1$ at epochs of a rate $1$ Poisson process, as long as the former site has strictly more particles than the latter. When $c=1$, this is the standard TASEP. Our main results address the asymptotics of the stationary distribution of a finite system, and especially the limit of the flux (current) as $N\to\infty$. In particular, we prove that interesting non-trivial phase transitions take place in a system with $c>1$. For example, if $c>1$ and $1/2 \le β\le 1$, the maximum limiting flux $1/4$ is achieved as long as $α\ge α_c^*$, where $α_c^* < 1/2$ is some non-trivial threshold. (For the standard TASEP the threshold is $1/2$.) We also put forward a general conjecture about the stationary distribution asymptotics under an arbitrary parameter setting. We illustrate our formal results and the conjecture by simulations, and identify interesting directions for further research.

Low-latency, resource-efficient neural network inference on FPGAs is essential for applications demanding real-time capability and low power. Lookup table (LUT)-based neural networks are a common solution, combining strong representational power with efficient FPGA implementation. In this work, we introduce KANELÉ, a framework that exploits the unique properties of Kolmogorov-Arnold Networks (KANs) for FPGA deployment. Unlike traditional multilayer perceptrons (MLPs), KANs employ learnable one-dimensional splines with fixed domains as edge activations, a structure naturally suited to discretization and efficient LUT mapping. We present the first systematic design flow for implementing KANs on FPGAs, co-optimizing training with quantization and pruning to enable compact, high-throughput, and low-latency KAN architectures. Our results demonstrate up to a 2700x speedup and orders of magnitude resource savings compared to prior KAN-on-FPGA approaches. Moreover, KANELÉ matches or surpasses other LUT-based architectures on widely used benchmarks, particularly for tasks involving symbolic or physical formulas, while balancing resource usage across FPGA hardware. Finally, we showcase the versatility of the framework by extending it to real-time, power-efficient control systems.

We construct proper good moduli spaces for moduli stacks of Bridgeland semistable orthosymplectic complexes on a complex smooth projective variety, which we propose as a candidate for compactifying moduli spaces of principal bundles for the orthogonal and symplectic groups. We also prove some results on good moduli spaces of fixed point stacks and mapping stacks from finite groupoids.

Optically pumped atomic magnetometers (OPAMs) offer high sensitivity at room temperature and are increasingly considered for portable magnetic sensing in geomagnetic-field environments. Here we report a handheld-scale, single-beam scalar $^{87}$Rb OPAM with a sensor-head volume of approximately 110~mL. The device operates in an all-optical Bell-Bloom configuration and uses digital lock-in, dispersive tracking of the $^{87}$Rb Larmor resonance, implemented with a hybrid electronics stack that combines in-house control hardware with commercial modules. A single frequency-modulated laser beam performs both pumping and probing without RF coils. All signal processing is realized in Python on a single-board computer paired with a commercial off-the-shelf (COTS) data-acquisition module, enabling immediate deployment without dedicated signal-processing hardware. The magnetometer has an intrinsic in-band field sensitivity of approximately 21~pT/$\sqrt{\mathrm{Hz}}$, estimated from the lock-in dispersion slope, over a 0.1--30~Hz closed-loop in-band region with a digital-output rate of 200~samples/s. In an unshielded Earth-field deployment, we detect repeatable transient magnetic signatures from a controlled elevator motion sequence and quantify standoff observability over sensor-elevator distances from 1.25~m to 10~m. These results show that compact scalar OPAMs can provide bandwidth and range-resolved event sensitivity suitable for field-deployable magnetic anomaly detection and infrastructure monitoring in realistic geomagnetic environments.

We establish a sharp Adams-type inequality in higher-order function spaces with singular weights on $\mathbb{R}^n$. A sharp singular concentration-compactness principle, improving Lions' result, is also proved. The study distinguishes between critical and subcritical sharp singular Adams-type inequalities and shows their equivalence. Furthermore, we analyze the asymptotic behavior of the associated bounds and relate the suprema of the critical and subcritical cases. A new compact embedding, crucial to our analysis, is also derived. Moreover, as an application of these results, by employing the mountain pass theorem, we study the existence of nontrivial solutions to a class of nonhomogeneous quasilinear elliptic equations involving the $(p,\frac{n}{2})$-biharmonic operator with singular exponential growth.